空间关联规则挖掘算法的研究与应用-软件工程专业论文.docx

摘要规则；算法用二进制数表示空间拓扑关系改进了数据存储方式，并用数字递增方式产生候选频繁项，实现空间拓扑关联挖掘，模拟实验表明了算法的高效性；将 其应用到电网可视化管理系统中，删除与停电操作不相关的设备，减少拓扑分析 搜索的节点或线元素数目，提高“最优化停电方案分析”功能的执行效率，通过系统 性能评估体现了算法的实用性。关键词：空间关联规则，交替搜索，横向挖掘，二进制，电网分析ABSTRACTABSTRACTSpatialdataminingextractslatentforecastinginformationandfindsthemost worthwhileknowledgetoguideScientificDecisionMakingfromspatialdatabase, whichisahottopicofresearchandapplicationforpeople.Atpresent,inspatial associationrulesmining,theminingmethodbasedonspatialtransactionisa comprehensiveappliedtechnologybypeople,butthetechnologyofformingfrequent itemsetsandpruningisoneofdifficultproblemswhenthesealgorithmsareappliedto huge spatialdata mining.Asfastdevelopmentofdigitizationelectricpowersystem,itisadisquisitive emphasisthatspatialdataminingisusedinelectricpowersystem.InGridVisual ManageSystem,theefficiencyofgridanalysisismainlyaffectedbythenumberof nodeandlinesearchedbytopologyanalysis.Becauseofshortcomingsofpresented miningalgorithms,whichinefficientlyimprovethespeedoftopologyanalysis,we needresearchsomeefficientalgorithmsofspatialassociationrulesmining,whichare used in GridVisual Manage Systemto improve efficiency of grid analysis.Inthispaper,aimingtotheshortageofspatialassociationtransversemining, namely,inpresentedthesespatialtransverseminingalgorithms,althoughthese algorithmsimprovethetechnologyofformingcandidatefrequentitemsetsandpruning, theyinefficientlyextractmonolayertransversespatialassociationrulesthatcontain morethenumberofspatialobject.Firstly,analgorithmofspatialtransactionmining basedonalternatesearch(ASTMAS)isproposed,whichissuitableforminingthese associationamongthesedifferentspatialobjectsfromthesamespatialpattern.Inhuge spatialdatamining,thealgorithmextractsmonolayertransversespatialassociation rulesthatcontainanythenumberofspatialobjects,throughchangingthetraditional wayofformingfrequentitemsetsandsearchstrategyofpresentedbinarymining algorithms.Thealgorithmusestwowaysofnumberascendinganddescendingto doublegeneratecandidatefrequentitemsets,ino